Econometric Reviews, a scholarly econometrics journal, has accepted for publication an article co-authored by Kostas Kyriakoulis, an academic consultant for The Financial Risk Group. Kyriakoulis, along with colleagues in the United Kingdom, presented “Inference in the Presence of Redundant Moment Conditions and the Impact of Government Health Expenditure on Health Outcomes in England.”
The paper, Kyriakoulis explained, compares two different estimation techniques: the Info-Metric estimation method (Empirical Likelihood and Exponential Tilting) and GMM (Generalized Method of Moments). Both models are estimated through a constrained optimization problem. Kyriakoulis wrote the code for both in Matlab.
“Both methods are very popular because they can estimate linear or non-linear models (single equations or systems) without making any distributional assumptions,” Kyriakoulis said. “Using a simulation study, we performed a comparative study of the finite performance of these estimators. The results indicate that the properties of GMM procedures deteriorate as the number of redundant/noisy information increases. “
In contrast, he said, the IM methods provide reliable point estimators in these cases. Any problems with confidence intervals and hypothesis testing can be corrected through bootstrap techniques, which Kyriakoulis and his colleagues performed.
To read the full paper, please visit https://econpapers.repec.org/paper/mansespap/1401.htm.
Econometric Reviews is a publication of Emory University. The journal specializes in articles and surveys of current economic conditions, and targets econometricians, business economists, statisticians, sociologists and psychologists.